To assess the effects of lipid-lowering therapy with lovastatin on coronary angiographic findings in patients with coronary artery disease and to compare the findings with those of two lipid-lowering angiographic trials using similar end points. Randomized, double-blind, placebo-controlled, multicenter coronary angiographic trial. Community- and university-based cardiac catheterization laboratories. A total of 270 patients, 37 to 67 years old, with total cholesterol ranging from 4.92 to 7.64 mmol/L (190 to 295 mg/dL) and angiographically defined coronary artery disease. A cholesterol-lowering diet and either lovastatin, 80 mg/day, or placebo. Per-patient change in percent diameter stenosis as determined by quantitative coronary angiography (primary end point). Global change score, based on the consensus of blinded expert readers regarding angiographic change (secondary endpoint). Lovastatin lowered total cholesterol level by 32%, low-density lipoprotein cholesterol by 38%, and the apolipoprotein B by 26% and raised the high-density lipoprotein cholesterol by 8.5% (P < 0.001). Average percent diameter stenosis increased 2.2% in placebo recipients and 1.6% in lovastatin recipients (P > 0.20). For lesions 50% or greater, average percent diameter stenosis increased 0.9% in placebo recipients and decreased 4.1% in lovastatin recipients (P = 0.005). The mean global change score was +0.9 (indicating progression) in the placebo group and +0.4 in the lovastatin group (P = 0.002); 13 placebo recipients and 28 lovastatin recipients had global change scores indicating regression (P < 0.02). Treatment with lovastatin plus diet slows the rate of progression and increases the frequency of regression in coronary artery lesions (by global change score), especially in more severe lesions (by quantitative angiography). This is the third lipid-lowering trial to show a benefit using the global change score, an end point predictive of clinical coronary events. Differences between two of these trials, using quantitative coronary angiographic end points, may have theoretical bearing on the mechanisms by which lipid-lowering therapy operates at the level of the arterial wall. * For current author affiliations and addresses, see end of text. See the appendix for a listing of the members of the MARS Research Group. Dr. Blankenhorn died on 9 May 1993.